1. IntroductionBlending of polymers is often an important routefor obtaining new and improved polymeric materi-als, which are difficult to obtain by direct polymer-ization process [1, 2]. In addition to the family ofsynthesized thermoplastic elastomers (TPEs),advances in the rubber-plastic blends lead to thedevelopment of new avenues to prepare TPEs withtailor made properties by varying the blend ratio ofthe components. Thermoplastic vulcanizates (TPVs)are special class of TPEs prepared by dynamic vul-canization process which involves the crosslinkingof rubber part while it is being mixed with themolten thermoplastic at high shear rate [3–5].Amongst the various blends, polypropylene (PP)/

ethylene propylene diene rubber (EPDM) blendshave gained tremendous importance due to com-mercial interest because of the structural compati-bility. A large number of reports have been pub-lished on the PP/EPDM uncrosslinked and dynami-cally crosslinked blends [3–5]. Recently, newfamilies of ethylene-alpha olefin copolymers havegained significant attention due to their uniquemolecular architecture with narrow molecularweight distribution, uniform co-monomer distribu-tion and long chain branching along the polymerbackbone [6]. Ethylene octene copolymer (EOC) inparticular has been shown to provide better tough-ening characteristics to PP, when compared withthe conventionally used EPDM rubber [7, 8].

197

*Corresponding author, e-mail: knaskar@rtc.iitkgp.ernet.in© BME-PT

Interrelationships of morphology, thermal and mechanicalproperties in uncrosslinked and dynamically crosslinkedPP/EOC and PP/EPDM blends

R. R. Babu, N. K. Singha and K. Naskar*

Rubber Technology Centre, Indian Institute of Technology, Kharagpur-721302, West Bengal, India

Received 22 December 2009; accepted in revised form 28 January 2010

Abstract. Thermoplastic vulcanizates (TPVs) based on polypropylene (PP) with ethylene octene copolymer (EOC) andethylene propylene diene rubber (EPDM) have been developed by coagent assisted dicumyl peroxide crosslinking system.The study was pursued to explore the influence of two dissimilar polyolefin polymers (EOC and EPDM) having differentmolecular architectures on the state and mode of dispersion of the blend components and their effects with special referenceto morphological, thermal and mechanical characteristics. The effects of dynamic crosslinking of the PP/EOC andPP/EPDM have been compared by varying the concentration of crosslinking agent and ratio of blend components. Theresults suggested that the uncrosslinked and dynamically crosslinked blends of PP/EOC exhibit superior mechanical prop-erties over PP/EPDM blends. From the hystersis experiments it was found that PP/EOC blends also perform better fatigueproperties over PP/EPDM based blends. It was demonstrated that, the origin of the improved mechanical properties of EOCbased blends is due to the combined effect of the unique molecular architecture with the presence of smaller crystals andbetter interfacial interaction of EOC phase with PP as supported by the results of thermal and fatigue analyses.

Keywords: polymer blends and alloys, polypropylene, ethylene octene copolymer, ethylene propylene diene rubber, ther-moplastic vulcanizate

eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209Available online at www.expresspolymlett.comDOI: 10.3144/expresspolymlett.2010.26

Presently, EOC is used for the impact modificationof PP for various applications especially in theautomotive sectors. Moreover, these materials areavailable in the form of pellets that help in easyhandling, faster mixing and compounding. Never-theless, reports on dynamically vulcanized PP/EOCTPVs are scanty. In case of TPVs, in addition to theconstituent polymers, type of curing agent plays amajor role in determining the final performance ofthe TPVs. Among several crosslinking systemsused in the preparation of PP/EPDM TPVs, pheno-lic resin gained more commercial success due to itscrosslinking and compatiblizing tendency [2].However, in this particular PP/EOC blend systemphenolic resin is ineffective, since the latter needsthe presence of unsaturation to form a crosslinkednetwork structure. Furthermore, use of phenolicresin in PP/EPDM blends causes black specks inthe products which motivates the development ofother crosslinking systems [9]. In that case, perox-ides can be used which crosslink both saturated andunsaturated polymers without any reversion char-acteristics. However, the activity of peroxidedepends on the type of polymer and presence ofother ingredients in the system. It has been wellestablished that PP exhibits β-chain scission(degradation) with the addition of peroxide [10].Hence the use of peroxide alone is limited in thepreparation of PP based TPVs. In general, to mini-mize the unwanted side reactions, coagent assistedperoxide crosslinking system is recommended.Naskar et al. [11, 12] studied the influence of dif-ferent peroxides and multifunctional peroxide (hav-ing peroxide and coagent functionality in a singlemolecule) in the conventional PP/EPDM blends.Recently, Naskar et al. [13] have been involved in

the development of a novel PP/EPDM TPVs byelectron beam induced reactive processing. Specialpurpose TPVs based on silicone rubber [14] andmaleated ethylene propylene rubber (m-EPM) [15]were also developed from the same lab. Lai et al.[16] investigated the fracture behavior and physico-mechanical properties of peroxide cured PP/EOCTPVs. As a part of the extensive research work onthe peroxide cured PP/EOC TPVs, our earlierreport deals with the influence of different perox-ides [17], different coagents [18] and different mix-ing sequences [19, 20] on the solid and melt stateviscoelastic properties of PP/EOC TPVs.The present study primarily aims to evaluate theperformance of the coagent assisted peroxide curedPP/EOC TPVs as a potential alternative to the con-ventional PP/EPDM TPVs. In this work, the mor-phological evolution and its influence on themechanical and thermal properties have beeninvestigated for both PP/EOC and PP/EPDMuncrosslinked and dynamically crosslinked blends.Detailed comparisons were pursued on morpholog-ical, thermal and mechanical properties. Investiga-tion had been made to correlate the observed resultswith the microstructural architecture of the con-stituent polymers.

2. Experiemental2.1. MaterialsIsotactic polypropylene (PP), ethylene-octenecopolymers (EOC) and ethylene propylene dienerubber (EPDM) were used as blend components.Metallocene catalyzed ethylene octene copolymers(Exact® 5171) were employed in this study. Theoctene content of these Exact copolymers was

198

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Table 1. Molecular characteristics of the polymers

MFI – Melt flow index; PDI – Polydispersity index; Mn – number average molecular weight; Mw – Weight average molecular weight*measured at a very low shear rate using Smart Rheo (Ceast, Italy) Capillary Rheometer at 180°C with 40:1 L/D capillary

Polymer/properties PP EOC EPDMSpecific gravity 0.90 0.87 0.86MFI [g/10 min] 3.0 at 230°C/2.16 kg 1.0 at 190°C/2.16 kg –ML (1+4) at 125°C – 25 55Mn [g/mol] 218 967 277 751 273 082Mw [g/mol] 949 319 494 872 622 876PDI (Mw/Mn) 4.33 1.78 2.28Octene content [%] – 13 –Ethylidene norbornene (ENB) content [%] – – 4.6Propylene content [%] 99.9 – 25.4Zero shear viscosity [Pa·s]* 1020 1980 4130Grade Koylene ADL (AS030N) Exact 5171 Keltan 5508Supplier IPCL, India ExxonMobil Chemical, USA DSM Elastomers, The Netherlands

determined from Proton Nuclear Magnetic Reso-nance Spectroscopy (1H-NMR) measurement. Thematerial and molecular characteristics of PP, EOCand EPDM are given in Table 1. The polymerstaken for comparison are EOC and EPDM whichwere selected on the basis of number averagemolecular weight that lies relatively in the samerange. Dicumyl peroxide (DCP) (Perkadox-BC-40B-PD) having active peroxide content of 40%;temperature at which half life time (t1/2) is 1 hour at138°C; specific gravity of 1.53 g/cm3 at 23°C wasobtained from Akzo Nobel Chemical Company,The Netherlands. N-N′-m-phenylene dimaleimide(MPDM) (SR 525; specific gravity, 1.44 g/cm3 at23°C) was obtained from Sartomer Company,USA.

2.2. Preparation of TPVs

TPVs were prepared by melt mixing of PP withEOC/EPDM (rubbery material) in a HaakeRheomix 600s with a mixing chamber volume of85 cm3 at a temperature of 180°C. Batch sizes wereabout 60 g. Total mixing time for each batch was14 minutes. At first, PP was allowed to soften for3 minutes and then rubbery material was added tothe chamber and melt mixing was carried out for5 minutes. Dynamic vulcanization was pursued byadding co-agent assisted peroxide for 2 minutes.Mixing was continued for another 4 minutes tocomplete the vulcanization. The compositions ofTPVs employed for this study are shown inTable 2. For better understanding, blend ratios aredesignated as X, Y and Z for 75, 50 and 25% of PPrespectively. TPVs prepared by different rubberymaterials are designated as follows: C for EOC andD for EPDM. For example, YC corresponds to thecomposition of 50:50 PP/EOC uncrosslinkedblends and ZD for 25:75 PP/EPDM uncrosslinkedblend. Dynamically vulcanized blends are desig-

nated with prefixes D2 or D4 followed by corre-sponding blend ratio designation. The letter D2 orD4 corresponds to the type and concentration ofperoxide (dicumyl peroxide of 2 phr or 4 phr). Forexample, D2ZC corresponds to the compositioncontaining 2 phr (parts per hundred grams of rub-ber) of dicumyl peroxide in the 25:75 PP/EOCblend ratio. After mixing, the blends were removedfrom the chamber at hot condition and sheeted outon a two roll mill in a single pass. Sheets were thencut and pressed in a compression molding machine(Moore Press, Birmingham, UK) at 190°C for4 min at 5 MPa pressure. Aluminum foils wereplaced between the mold plates. The molded sheetswere then cooled down to room temperature underthe same pressure.

2.3. Testing procedures

Morphology – Morphology studies were carried outusing a scanning electron microscope (SEM)(JEOL JSM 5800, Japan). Molded samples of PP/EOC or PP/EPDM TPVs were cryofractured in liq-uid nitrogen to avoid any possibility of phase defor-mation during the fracture process. In case ofuncrosslinked blends, the EOC or EPDM phasewas preferentially extracted by treating with hotxylene at 50°C for 15 min, whereas in dynamicallyvulcanized blends the PP phase was preferentiallyand partially extracted by etching with hot xyleneat 100°C for 45 min. The samples were then driedin a vacuum oven at 70°C for 5 hours to remove thetraces of solvents present. Treated surfaces werethen sputtered with gold before examination.Differential Scanning Calorimetry (DSC) – DSCmeasurements were carried out in a TA instrument,USA (Model DSC Q100 V 8.1) to study the melt-ing behavior. The samples of about 6 mg sealed inaluminum pans were heated from –80 to 200°C at ascanning rate of 10°C/min under nitrogen atmos-phere. The degree of crystallinity of PP was deter-mined from the heating curve by using standardformula which was discussed earlier by the authors[17, 18].Mechanical Testing – The dumb-bell shaped speci-mens of the TPV used for testing were die cut fromthe compression molded sheet and the testing wasdone after 24 hours of maturation at room tempera-ture. Tensile strength was measured according toASTM D418-98A using a universal testing

199

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Table 2. TPV compositions in phr

Prefix: D2 – Dynamically vulcanized blends with 2 phr DCPD4 – Dynamically vulcanized blends with 4 phr DCP

Polymers/Designations

PP EOC EPDM

XC 75 25 –YC 50 50 –ZC 25 75 –XD 75 – 25YD 50 – 50ZD 25 – 75

machine Hounsfield H10KS (UK) at a constantcross-head speed of 500 mm/min. Tension set wasperformed at room temperature with a stretchingcondition for 10 minutes at 100% elongationaccording to the ASTM D412-98 method.Cyclic Tensile Testing – The tensile testing speci-men was used to perform the hysteresis and fatiguelife testing of the samples. The loading and unload-ing cross-head speed of 500 mm/min was main-tained. All the values were measured at the roomtemperature (25°C) in air without any pretreatment.The hysteresis measurements were done in twomodes i.e. constant strain amplitude and stepwiseincrease in strain amplitude.Overall Crosslink Density (OCD) – Equilibriumsolvent swelling measurements were carried out onthe PP/EOC TPVs to determine the crosslink den-sity of the EOC (ν) in presence of PP [11, 17]. Theoverall crosslink density was calculated using themodified Flory–Rehner equation [21]. From thedegree of swelling, the overall crosslink densitywas calculated relative to the (EOC+PP) phasesand expressed as (ν+PP). The latter was done inorder to avoid the need to correct for a part of thePP, being extracted as amorphous PP. A circularpiece of 2 mm thickness was made to swell incyclohexane for about 48 hours to achieve equilib-rium swelling condition. Initial weight, swollenweight and de-swollen or dried weight were meas-ured and substituted in the Equation (1):

(1)

where ν – number of moles of effectively elasticchains per unit volume of EOC [mol/ml] (crosslinkdensity), (ν+PP) – crosslink density of EOC phasein the presence of PP (overall crosslink density),

Vs – molar volume of cyclohexane [cm3/mol], χ –polymer swelling agent interaction parameter,which in this case is 0.306 [11] and Vr – volumefraction of ethylene-octene copolymer in theswollen network, which can be expressed by Equa-tion (2):

(2)

where Ar is the ratio of the volume of absorbedcyclohexane to that of ethylene-octene copolymerafter swelling.

3. Results and discussion3.1. Processing characteristicsMixing torque-time relationship (processing char-acteristics) of uncrosslinked and dynamicallycrosslinked blends were recorded from HaakeRheomix. The stabilized torque value is the charac-teristic feature of the viscous nature of the melt. Itclearly indicates that the stabilized torque valueincreases with increase in rubber (EOC or EPDM)content as shown in Table 3. As mentioned previ-ously (Table 1), EPDM has higher molecularweight and chain entanglement density relative toEOC. Therefore, EPDM macromolecules showhigher resistance to the chain conformation andthereby show higher torque values than EOC basedblends. In the dynamically vulcanized blends, theprocessing characteristic is more or less similar tothat of corresponding uncrosslinked blends untilthe curative (peroxide with coagent) is added.When the curative is added, torque value increases,reaches maximum and then decreases before beingstabilized. Typical torque-time relationship of50/50 PP/EOC and PP/EPDM uncrosslinked and

1

1

+=

rr A

V

rr

rrr

s VV

VVV

VPP

5.0)(

)()1ln(·

1)(

3/1

2

−χ++−−=+ν

200

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Table 3. Mixing torque values of the uncrosslinked and dynamically crosslinked blends

Compoundname

PP/EOCCompound

name

PP/EPDMMaximum torquevalue after adding

curative [N·m]

Torque value at theend of the mixing

cycle [N·m]

Maximum torquevalue after adding

curative [N·m]

Torque value at theend of the mixing

cycle [N·m]XC – 3.5 XD – 4.0YC – 5.3 YD – 6.8ZC – 6.8 ZD – 11.4

D2XC 7.6 2.5 D2XD 6.9 1.8D2YC 14.0 7.4 D2YD 14.3 7.2D2ZC 20.6 17.3 D2ZD 22.7 017.2D4XC 7.4 1.6 D4XD 7.0 1.2D4YC 15.2 7.0 D4YD 15.5 6.8D4ZC 23.3 20.1 D4ZD 25.5 19.8

dynamically crosslinked blends are shown in Fig-ure 1. The increase in torque is due to the crosslink-ing occurring in the EOC or EPDM phases.Decrease in torque value after reaching the maxi-mum is expected due to the combined effect of thedisintegration of crosslinked EOC or EPDMdomains and also by the degradation in the PPphase through β-scission [22]. The maximumtorque value after adding the curative and thetorque value at the end of the mixing cycle for dif-ferent compositions are given in the Table 3.Depending on the composition and polymer type,extent of degradation and crosslinking might beresolved. Samples with higher amount of PP (XCand XD) having 75% of PP, suffer severe degrada-tion in PP phase during dynamic vulcanization.Therefore end torque (final torque) value decreaseswith the addition of curative with respect touncrosslinked blends. Nevertheless, torque valueincreases initially after the addition of curative,which may be associated with the crosslinking thattakes place in the EOC or EPDM phase. In case ofYC and YD blends containing 50% of PP, thereexists equal contribution to crosslinking in the rub-bery phase and chain-scission in the PP phase. It isworth noting that the maximum torque increasesand end torque decreases as a function of curativedosage. In ZC and ZD blends containing 25% of PPcurative is mostly consumed by the rubbery phaseand overrides the degradation in PP phase. Thisreflected to increase both the end torque and maxi-mum torque values as a function of curative dosage(within the curative concentration limit). The aboveresults suggest that the maximum torque value

increases with increase in rubber content in both thesystems (PP/EOC and PP/EPDM).It is to be noted here that, EPDM based TPVs showlower end torque values than corresponding EOCbased TPVs at identical rubber/plastic composition.EPDM is a random terpolymer of ethylene, propy-lene and diene units and with the addition of perox-ide, propylene units undergo chain-scission and/ordisproportionation as side reaction apart from thecrosslinking in the ethylene units. Hence EPDMbased TPVs suffer severe degradation than EOCbased TPVs. Apparently, peroxide at the highercurative dosage (4 phr) not only increases theextent of crosslinking in EOC or EPDM phase butalso causes significant degradation in the PP phase.

3.2. Morphology

In general, phase morphology of the heterogeneouspolymer blends are mainly determined by viscosityratio, blend ratio and interfacial tension [23]. Sev-eral authors [24–27] surveyed the morphology andrheology of various kinds of polyolefin blends withspecial reference to the varying blend ratios.Figure 2 shows the SEM photomicrographs of50/50 blend ratio of PP/EOC and PP/EPDMuncrosslinked and dynamically crosslinked blends.In uncrosslinked blends, rubbery phase EOC orEPDM was preferentially extracted and in dynami-cally crosslinked blends PP phase was preferen-tially extracted. Both the uncrosslinked compound(YC and YD) results in the formation of co-contin-uous phase morphology. Depending on the viscos-ity ratio, various levels of co-continuity exist in theblend compounds. It is clear from the Figures 2aand 2b that the PP/EOC (YC) contributes to havehigher percentage of co-continuity than PP/EPDM(YD) blend. Because of higher viscosity ratio inYD, EPDM prefers to exist in dispersed phase.However by dynamic vulcanization, a completephase transition (inversion) occurs i.e. co-continu-ous morphology is changed into dispersed phasemorphology. The less viscous PP encapsulates themore viscous crosslinked rubber particles to mini-mize the mixing energy [22, 23]. At equal curativedosage of 2 phr DCP and compositional volume,PP/EPDM TPV shows coarser crosslinked rubberparticle size than PP/EOC TPV (Figures 2c and2d). By increasing the concentration of peroxide(4 phr) in order to increase the crosslink density of

201

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Figure1. Comparative mixing torque-time relationship of50/50 PP/EOC and PP/EPDM uncrosslinked anddynamically crosslinked blends

rubber phase, two opposite behaviors have beenaccounted in the particle size evolution for PP/EOCand PP/EPDM blend. Crosslinked rubber particlessize increases in PP/EOC blends, whereas decreasesin PP/EPDM blend as a function of peroxide con-centration (Figures 2e and 2f). It is well knownthat, PP undergoes chain-scission with the additionof peroxide. In case of PP/EOC blends when theconcentration of peroxide increases from 2 to 4 phr,extent of degradation in the PP phase dominatesover crosslinking in the EOC phase. Therefore, itbecomes too difficult for a relatively very low vis-cosity PP matrix (low molecular weight PP) tobreak-up the crosslinked EOC rubber. But in caseof PP/EPDM blends, in addition to the PP-matrixchain-scission, the inherent PP segments in theEPDM also undergo chain-scission which cause toform low molecular weight compound. These lowmolecular weight compounds have higher tendencytowards fine dispersion by lowering the viscosityratio.

3.3. Thermal properties

DSC measurements were performed to characterizethe crystallization and melting behavior ofuncrosslinked and dynamically vulcanized blends.Figure 3 shows the DSC cooling and heating curvesof 50/50 rubber/plastic blend ratio of uncrosslinkedand dynamically vulcanized blends. Table 4 sum-marizes the relevant thermal parameters as obtained

from the DSC scans. It can be seen from theTable 4, that neat PP exhibits a characteristic fea-ture of semi-crystalline polymer with glass transi-tion (Tg), melting temperature (Tm) and crystalliza-tion temperature (Tc) at around 1, 165 and 119°C,respectively. Pristine EOC shows a strong β-transi-tion (step in DSC heat capacity traces) due to itsmovement of octene branch points at –50°C. Oncontrary, EPDM shows a transition at –45°C due tothe relaxation of short chain branch points. But, PPdoes not show a significant change in β-transitionin the DSC traces due to its semi-crystalline charac-teristics. Both EOC and EPDM show the occur-

202

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Figure 3. DSC curves of melting endotherm and crystal-lization exotherm of 50/50 rubber/plasticuncrosslinked and dynamically crosslinkedblends. (second heating scan and cooling scan ofthe blends after subjecting them to same thermalhistory)

Figure 2. SEM photomicrographs of a) YC (50/50 PP/EOC uncrosslinked) b) YD (50/50 PP/EPDM uncrosslinked)c) D2YC (dynamically vulcanized; 2 phr) d) D2YD (dynamically vulcanized; 2 phr) e) D4YC (dynamicallyvulcanized; 4 phr) f) D4YD (dynamically vulcanized; 4 phr)

rence of Tm and Tc, suggesting the presence of crys-talline phase. It is expected that, both EOC andEPDM seem to have long enough ethylene blocks,which are easy to crystallize in a usual manner.However, depending on its inherent moleculararchitecture Tc of EOC and EPDM are 45 and 20°Crespectively.At 50/50 rubber/plastic blend ratio (YC and YD),Tm, Tc and Xc found to vary significantly for bothPP as well as for rubbery component. The possibleinter-diffusion and inter-molecular mixing reducesthe size and perfectness of crystals and conse-quently inhibits the crystallization process on eithercomponent as also reported by Bielinski et al. [28].The results, therefore, suggest that PP blended withEOC or EPDM needs higher undercooling to crys-tallize (i.e., rubbery components may depress thecrystallization of PP) and correspondingly decreasethe Tc. It is worth noting that the addition of rub-bery component broadens the melting endothermsof PP phase, which is probably due to the changesin the distribution of crystal lamella thickness.Upon dynamic vulcanization, significant changesin the thermal properties were observed in both PPand rubbery phases. Apparently, as the peroxidedosage increases from 0 to 4 phr in PP/EOC blend,Tg marginally increases, Tm as well as Xc decreasecontinuously in the EOC phase due to low degreeof chain alignment resulting from predominantcrosslinking. As for PP matrix phase, variation inthe Tg is marginal whereas Tm and Xc continuouslydecrease as the curative concentration increases. Asimilar trend is found for the PP/EPDM TPVs. It isvery clear from the Table 4, the extent of decreasein Xc of PP is more pronounced for PP/EOC TPVsas a function of curative concentration i.e., Xc of PPdecreased from 46.6 to 32.4% with addition of

4 phr of peroxide. Bulk crystallization temperaturesof both the components were found to vary signifi-cantly by dynamic vulcanization. As for the rub-bery phase, Tc(EOC) was found to decrease continu-ously in PP/EOC TPVs. This may be associatedwith the prominent crosslinking reaction in theEOC phase which needs higher under-cooling tocrystallize. Whereas in case of PP/EPDM TPVs,EPDM undergoes simultaneously crosslinking inthe ethylene sequence and chain-scission in thepropylene sequence. The combined effect of theabove mentioned two competing reactions incon-sistently affect the crystallization process of EPDMphase. As the peroxide dosage increases from 0 to4 phr (with in the given concentration limit), theTc(PP) was found to increase and then decrease forPP/EOC based TPVs. At 4 phr concentration ofperoxide, chain scission in the PP phase dominatesover the crosslinking effect in the EOC phase.Therefore, the lower molecular weight PP chainshave less under-cooling effect to crystallize. But incase of PP/EPDM TPVs, Tc(PP) initially increasesand then gets stabilized, which is mainly due to thecompetition between the chain coupling reaction inthe polyethylene segments and chain scission reac-tion in the polypropylene segments in the EPDMphase.

3.4. Mechanical properties

The effect of blend composition on the tensileproperties of PP/EOC and PP/EPDM uncrosslinkedand dynamically vulcanized blends are comparedand shown in Table 5. Comparative stress-strainproperties of the blend components are shown inFigures 4a and 4b. Neat PP shows very high tensilestrength and poor elongation, a characteristic fea-

203

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Table 4. Crystallization and melting characteristics of uncrosslinked and dynamically crosslinked blends

Tg – Glass transition temperature, Tm – Peak melting temperature, ΔHf – Heat of fusion value, Xc – Percentage crystallinity calculatedfrom heat of fusion data, Tc – Crystallization temperature

Compoundname

PP EOC/EPDMTg [°C] Tm [°C] ΔΔHf [J/g] Xc [%] Tc [°C] Tg [°C] Tm [°C] ΔΔHf [J/g] Xc [%] Tc [°C]

PP 1 165 99.32 47.52 119 – – – – –EOC – – – – – –50 60 18.9 6.55 45

EPDM – – – – – –45 38 08.2 2.85 20YC 1 162 48.7 46.63 113 –47 60 07.5 5.17 42

D2YC 2 161 46.5 44.52 124 –46 58 06.4 4.41 40D4YC 2 160 33.9 32.44 120 –45 58 05.5 3.76 37

YD 0 162 45.8 43.82 114 –43 37 03.6 2.48 19D2YD 2 161 45.3 43.34 125 –38 33 05.1 3.51 18D4YD 2 158 45.2 43.22 124 –38 33 04.0 2.75 18

ture of a brittle material. On the other hand, EPDMbeing primarily an amorphous polymer havingunreliable amount of crystals due to the crystalliz-ing nature of long chain polyethylene segmentsshows a typical uncrosslinked elastomer like defor-mation. The stress-strain behavior of EOC is simi-lar to that of uncrosslinked rubber; however stressgradually increases with increase in strain level.This may be due to the crystallizing tendency ofpolyethylene segments during stretching (Fig-ure 4a, inset). A significant difference was observedin the deformation characteristics of the blends

with varying compositions (Figure 4a). SamplesXC and XD show necking and yielding region. Itmay be associated with the distribution and disper-sion of rubber particles in inter- and intra-spherulitic region of PP matrix to reduce the notchsensitivity and brittle fracture [29]. The yield stressvalue and the peak broadening are observed whenthe rubber content increases from 25 to 50% (YCand YD). In case of ZC and ZD, elongation at breakincrease and modulus and tensile strength increaseas compared to YC and YD. The improved rubberybehavior can be explained in terms of morphologi-

204

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Figure 4. Stress-strain properties of (a) Uncrosslinked blends Inset: Pristine polymer (b) Dynamically vulcanized blendswith 2 phr concentrations of peroxide Inset: 4 phr peroxide

Table 5. Comparative physical properties of uncrosslinked and dynamically crosslinked blends

aStandard deviation of overall crosslink density value lies in between 0.1–0.15

Compoundname

Tensile strength[MPa]

Elongation at break[%]

100% modulus[MPa]

Tension set[%]

Overall crosslink density(νν+PP)·10–4 [mol/ml]a

PP 39.6±1.2 0016±3 – – –EOC 14.4±0.8 1265±24 2.1±0.2 4±0.2 0.06

EPDM 3.51±0.4 1485±35 0.6±0.1 2.5±0.3 0.19XC 22.9±2.1 0091±11 – – 40.05YC 13.2±2.3 0155±15 12.9±0.5 36±1.2 8.17ZC 09.2±1.2 0795±21 5.4±0.4 14±0.6 0.38

D2XC 20.1±2.4 0016±3 – – 36.10D2YC 15.4±1.5 0237±12 14.0±0.3 32±0.8 10.12D2ZC 11.5±1.1 0300±15 6.8±0.2 12±0.3 2.58D4XC 18.3±2.4 0014±2 – – 32.21D4YC 14.5±1.2 0200±8 13.8±0.3 28±0.5 10.32D4ZC 11.5±0.5 0295±12 7.0±0.2 10±0.2 3.72

XD 19.7±1.6 0105±10 – – 46.15YD 11.9±1.0 0117±14 11.9±0.3 28±0.6 13.92ZD 04.5±0.4 0477±18 3.1±0.1 8±0.2 2.91

D2XD 19.5±2.5 0017±2 – – 39.21D2YD 12.5±1.3 0150±13 12.7±0.2 24±0.4 16.84D2ZD 09.9±0.8 0269±15 5.3±0.2 6±0.3 4.92D4XD 15.4±1.5 0093±9 – – 32.38D4YD 13.9±0.0 0160±16 12.5±0.3 22±0.5 16.21D4ZD 09.3±0.0 0222±19 5.7±0.2 5±0.2 5.94

cal evolution (phase transition of EOC or EPDMphase form co-continuous to continuous as the rub-bery content increases from 50 to 75% as shown inthe earlier sections).The stress-strain curve of dynamically vulcanizedblend with 2 phr concentration of peroxide isshown in Figure 4b. For the compositions D2XCand D2XD, degradation in the PP phase results toshow lower stress and strain values. In case ofD2YC and D2YD, the nature of the curve signifi-cantly changes from plastic to elastic type with thedisappearance of necking and yielding. Thereby,dynamically vulcanized samples exhibit improvedtensile strength and elongation at break. The abovechange in the deformation mechanism is mainlyassociated with the evolution of phase morphologyduring dynamic vulcanization. Samples with higherrubber content (D2ZC and D2ZD) are more towardsthe elastic type with increased modulus and tensilestrength and lower elongation at break. Further-more, at higher peroxide dosage (4 phr), D4XC andD4XD show poor tensile and elongation at breakvalues, which may be due to extensive degradationor chain-scission in the PP phase. In case of dynam-ically vulcanized 50/50 rubber/plastic content(D4YC and D4YD), marginal decrease in both ten-sile and elongation values suggest that, chain-scis-sion in PP dominates over crosslinking in EOC orEPDM phase. Interestingly, not much difference isobserved in the stress-strain values with increase inthe curative dosage for samples with higher rubbercontent (ZC and ZD) (Figure 4b, inset graph).When comparing the behavior of PP/EOC andPP/EPDM samples, in all the cases EOC baseduncrosslinked and dynamically vulcanized blendsexhibit better physical properties than PP/EPDMsamples at identical composition. Since EPDM isan ethylene propylene diene random terpolymer,part of the propylene units undergoes chain-scis-sion or degradation while using peroxide as cura-tive. During dynamic vulcanization, as the concen-tration of curative increases from 0 to 4 phr, ahigher level of degradation in the PP phase is cou-pled with the chain-scission in the PP segments inthe EPDM polymer in PP/EPDM blends. TherebyPP/EPDM TPVs lose their mechanical strengthdespite containing the finest vulcanized rubber par-ticles. Therefore, it is expected that dynamicallyvulcanized blends of PP/EPDM depict inferiorphysico-mechanical properties than PP/EOC TPVs.

From the DSC and SEM analyses, it was found thatEOC containing small crystals have good affinityand finely dispersed particles in the PP matrixphase. On the other hand, EPDM itself containssmall crystals but its crystallinity might be too low(due to high degree of chain entanglement density)to form an effective reinforcing characteristics.Although mechanical properties are found to besuperior for PP/EOC blends, set properties are bet-ter for PP/EPDM blends.Tension set and overall crosslink density (OCD)values show a good correlation for both PP/EOCand PP/EPDM uncrosslinked and dynamicallycrosslinked blends. As the concentration of perox-ide increases form 0 to 4 phr (in both the cases),samples with higher amount of PP suffer fromchain scission and hence OCD value decreases con-tinuously. However samples with 50 and 75% rub-ber content, OCD continuously increases. Particu-larly, the latter (75% rubber content) shows signifi-cant increase in the OCD values with increase inperoxide concentration. These changes are associ-ated with the chain coupling reactions in the rub-bery phase. As a mater of fact, EPDM based blendsshow higher OCD values as compared to EOCbased compounds. In case of EPDM based sam-ples, the presence of natural physical entangle-ments at their characteristics network densitywould cause an increase in number of effectivecrosslink points per molecule (crosslink density)and correspondingly improve the set properties.

3.5. Hysteresis

In general, hysteresis measurements give informa-tion about the structural changes of the materialunder cyclic deformation. The samples are sub-jected to uniaxial tensile deformation at a constantspeed of 500 mm/min and the deformation wasremoved at the same speed. In general, due to theviscoelastic nature of the polymeric materials, theenergy applied may not be balanced with energyreleased i.e., the stretching and recoiling curvesform a closed loop called as hysteresis loop. Thehysteresis loop of cyclic tensile curves of 50/50rubber/plastic blend ratio of uncrosslinked anddynamically crosslinked blends at constant strain of100% are shown in Figures 5a and 5b. Duringstretching, a 50:50 blend ratio of PP and rubberycomponents exhibit the stress-strain behavior, a

205

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

characteristic feature of a thermoplastic material:an initial linear elastic behavior followed by grad-ual yielding [30]. Compared to uncrosslinkedblends, TPVs exhibit steady strain hardening effectimmediately after yielding. According to Solimanet al. [31] the deformation mechanism of PP/EPDM TPVs (as shown in Figure 6) is the combi-nation of matrix yielding and buckling of yieldedPP matrix by the recovering tendency of crosslinkedrubbery particles. During stretching, PP acts as gluebetween the crosslinked EPDM particles and partof the PP fractions undergo yielding in the directionperpendicular to the load. During recovery, theyielded PP fraction is partially pulled back by therecovering tendency of the crosslinked particles.During recoiling, the residual strain or permanent

set is observed due to matrix yielding. All the sam-ples exhibit a decrease in the maximum stress levelwhich can be explained by strain softening effect. Itis clear from the Table 6 and Figure 5; significantsoftening is occurred in the first cycle followed bygradual softening effect with increase in number ofcycles. A plausible mechanism for the observedchanges could be due to disentanglement and slip-page of polymer chains as a consequence of matrixyielding or shear banding. In both the cases, (PP/EOC and PP/EPDM) the dynamically vulcanizedblends decrease the setting behavior and lower theenergy loss, a characteristics feature of elastomericmaterial. Furthermore, comparing the hysteresisbehavior of two material classes, D2YC andD2YD, it is seen that latter exhibits less settingbehavior in combination with lower energy loss. Itmay be due to the inherent physical crosslinks byhigher chain entanglement density.Samples have been subjected to stepwise increas-ing strain test (SIST) to understand the structuralchanges of the material as a function of cyclicfatigue straining. Figures 7a and 7b demonstratethe results of the SIST of uncrosslinked anddynamically crosslinked blends subjected to strain-ing at predetermined values of 20, 50, 100, 150 and180% at a same speed of 500 mm/min. From the

206

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Figure 5. Stress-strain hysteresis loops during cyclic tension at constant strain of 100% (a) 50/50 PP/EOC uncrosslinkedand dynamically crosslinked blends (b) 50/50 PP/EPDM uncrosslinked and dynamically crosslinked blends

Figure 6. Deformation mechanism of TPVs suggested bySoliman et al. [31]

Table 6. Results of hysteresis test at constant strain of 100%

Compoundname

Residual strain [%] Max. stress [MPa]1st cycle 2nd cycle 5th cycle 1st cycle 2nd cycle 5th cycle

YC 41.0 43.0 51.1 11.8 11.2 10.8YD 35.0 37.5 44.2 09.4 08.7 08.0

D2YC 34.3 36.4 43.3 11.8 10.9 10.2D2YD 27.8 29.5 29.0 11.3 10.3 10.1

Figures 7a and 7b, it is clear that the residual strainor irreversible strain is smaller in low strain regionbut becomes significantly higher at higher strainlevel. Therefore, the TPVs exhibit rubber like elas-tic deformation in the low strain region. But withincreasing the strain level, elastic deformation issuperimposed with irreversible plastic deformation.Although, PP/EPDM based uncrosslinked blendand dynamically blends exhibit lower setting ten-dency and energy loss than corresponding PP/EOCbased blends but with poor fatigue resistance (failedbefore completing fourth cycle). The improvedfatigue behavior of D2YC can be explained by thefact of better interfacial adhesion between the blendcomponents.

4. Conclusions

The morphological, thermal and mechanical resultsare interpreted in terms of the balance between thecrosslinking in the EOC and EPDM phases versusdegradation in the PP phase. It is worth noting that,PP segments in the EPDM rubber might alsoundergo chain-scission and/or disproportionationreaction during dynamic vulcanization. So theobtained results are indeed the balance of the abovementioned reactions in the constituent polymers.The morphology of PP/EOC and PP/EPDM blendsindicate a two phase morphology in which the rub-ber phase is dispersed in continuous PP matrix atlower rubber/plastic blend ratio and followed byphase inversion with increase in rubber/plasticblend ratio. The 50/50 composition of rubber/plas-tic shows co-continuous phase morphology. How-

ever, after dynamic vulcanization all the samplesexhibit a dispersed phase morphology in whichcrosslinked rubber particles are dispersed in contin-uous PP matrix. The dispersion characteristics ofthe dispersed crosslinked rubber phase are mainlyinfluenced by the melt viscosity ratio, shear rate,curative dosage and the rubber content.The crystallinity of the PP phase in TPVs is more orless comparable for both the systems employed.Therefore the stress-strain response of the PP/EOCand PP/EPDM dynamically crosslinked blends atidentical composition is found to depend on the dis-persion characteristics of the crosslinked rubberphase. Dynamically vulcanized blends havinghigher rubber/plastic ratio show thermoplastic elas-tomeric characteristics, whereas those havinghigher plastic/rubber ratio show poor mechanicalproperties due to degradation in the PP phase.When compared to the conventional PP/EPDMTPVs, PP/EOC TPVs show lower viscosity andelasticity in the melt state and better physico-mechanical properties in the solid state. Hystersismeasurements have been found to be a useful toolto evaluate the structural changes during dynamicdeformation and fatigue analysis. Dynamically vul-canized 50/50 rubber/ plastic blend shows lowerenergy loss and better set properties. Although,PP/EPDM TPVs show lesser relaxation than corre-sponding PP/EOC TPVs but show poor fatiguecharacteristics.In short, PP/EOC shows better mechanical, andthermal properties which may be due to better inter-facial interaction and unique molecular microstruc-ture of the EOC phase. In addition to lower cost,

207

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

Figure 7. Stepwise hysteresis loops during cyclic tension (a) 50/50 PP/EOC uncrosslinked and dynamically crosslinkedblends (b) 50/50 PP/EPDM uncrosslinked and dynamically crosslinked blends

EOC are supplied in free pellets which allow forboth batch and continuous production of TPVs. So,peroxide cured PP/EOC TPVs combine better pro-cessing characteristics with superior performancein the solid state. Hence, peroxide cured PP/EOCcan be considered as a potential alternative to theconventional PP/EPDM TPVs.

AcknowledgementsThe authors are grateful to the Council of Scientific andIndustrial Research (CSIR), New Delhi, India for financialassistance.

References[1] Paul D. R., Bucknall C. B.: Polymer blends. Wiley,

New York (2000).[2] Karger-Kocsis J.: Thermoplastic rubbers via dynamic

vulcanization. in ‘Polymer blends and alloys’ (eds.:Shonaike G. O., Simon G. P.) Marcel Dekker, NewYork (1999).

[3] De S. K., Bhowmick A. K.: Thermoplastic elastomersfrom rubber plastic blends. Horwood, London (1990).

[4] Coran A Y., Patel R., Williams D.: Rubber-thermo-plastic compositions. Part V. Selecting polymers forthermoplastic vulcanizates. Rubber Chemistry andTechnology, 55, 116–136 (1982).

[5] Naskar K.: Thermoplastic elastomer based on PP/EPDM blends by dynamic vulcanization – Review.Rubber Chemistry and Technology, 80, 504–510(2007).

[6] Walton K. L.: Metallocene catalysed ethylene/alphaolefin copolymers used in thermoplastic elastomers.Rubber Chemistry and Technology, 77, 552–568(2004).

[7] McNally T., McShane P., Nally G. M., Murphy W. R.,Cook M., Miller A.: Rheology, phase morphology,mechanical, impact and thermal properties ofpolypropylene/metallocene catalysed ethylene 1-octene copolymer belnds. Polymer, 43, 3785–3793(2002).DOI: 10.1016/S0032-3861(02)00170-2

[8] Da Silva A. L. N., Rocha M. C. G., Coutinho F. M. B.,Bretas R., Scuracchio C. J.: Rheological, mechanical,thermal, and morphological properties of polypropy-lene/ethylene-octene copolymer blends. Journal ofApplied Polymer Science, 75, 692–704 (2000).DOI: 10.1002/(SICI)1097-4628(20000131)75:5<692

::AID-APP12>3.0.CO;2-Y[9] van Duin M.: Recent developments for EPDM-based

thermoplastic vulcanisates. Macromolecular Sym-posia, 233, 11–16 (2006).DOI: 10.1002/masy.200690006

[10] Dluzneski P. R.: Peroxide vulcanization of elastomers.Rubber Chemistry and Technology, 74, 451–490(2001).

[11] Naskar K., Noordermeer J. W. M.: Dynamically vul-canized PP/EPDM blends: Effects of different types ofperoxide on the properties. Rubber Chemistry andTechnology, 76, 1001–1018 (2003).

[12] Naskar K., Noordermeer J. W. M.: Dynamically vul-canized PP/EPDM blends: Effects of multifunctionalperoxides as crosslinking agents. Rubber Chemistryand Technology, 77, 955–971 (2004).

[13] Naskar K., Gohs U., Wagenknecht U., Heinrich G.:PP-EPDM thermoplastic vulcanisates (TPVs) by elec-tron induced reactive processing. Express PolymerLetter, 3, 677–683 (2009).DOI: 10.3144/expresspolymlett.2009.85

[14] Basuli U., Chaki T. K., Naskar K.: Influence ofEngage® copolymer type on the properties of engage/silicone rubber-based thermoplastic dynamic vulcan-izates. Express Polymer Letters, 2, 846–854 (2008).DOI: 10.3144/expresspolymlett.2008.99

[15] Naskar K., Chatterjee K.: Development of thermo-plastic elastomers based on maleated ethylene propy-lene rubber and polypropylene by dynamic vulcaniza-tion. Express Polymer Letters, 1, 527–534 (2007).DOI: 10.3144/expresspolymlett.2007.75

[16] Lai S-M., Chin F-C., and Chiu T-Y.: Fracture behav-iors of PP/mPE thermoplastic vulcanizate via peroxidecrosslinking. European Polymer Journal, 41, 3031–3041 (2005).DOI: 10.1016/j.eurpolymj.2005.06.003

[17] Babu R. R., Singha N. K., Naskar K.: Studies on theinfluence of structurally different peroxides inpolypropylene/ethylene alpha olefin thermoplasticvulcanizates (TPVs). Express Polymer Letters, 2,226–236 (2008).DOI: 10.3144/expresspolymlett.2008.27

[18] Babu R. R., Singha N. K., Naskar K.: Dynamicallyvulcanized blends of polypropylene and ethyleneoctene copolymer: Influence of various coagents onmechanical and morphological characteristics. Journalof Applied Polymer Science, 113, 3207–3221 (2009).DOI: 10.1002/app.30000

[19] Babu R. R., Singha N. K., Naskar K.: Melt viscoelas-tic properties of peroxide cured polypropylene-ethyl-ene octene copolymer thermoplastic vulcanizates.Polymer Engineering and Science, 50, 455–467(2010).DOI: 10.1002/pen.21553

[20] Babu R. R., Singha N. K., Naskar K.: Effects of mix-ing sequence on peroxide cured polypropylene (PP)/ethylene octene copolymer (EOC) thermoplastic vul-canizates (TPVs). Part I. Morphological, mechanicaland thermal properties Journal of Polymer Research,in press (2010).DOI: 10.1007/s10965-009-9354-z

208

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209

[21] Flory P. J., Rehner J. J.: Statistical mechanics of cross-linked polymer networks. II. Swelling. Journal ofChemical Physics, 1, 521–526 (1943).DOI: 10.1063/1.1723792

[22] Varghese S., Alex S., Kuriakose B.: Natural rubber-isotactic polypropylene thermoplastic blends. Journalof Applied Polymer Science, 92, 2063–2068 (2004).DOI: 10.1002/app.20077

[23] Harrats S., Thomas S., Groeninckx G.: Micro andnano structured multiphase polymer blend systems:Phase morphology and interface. CRC Press, BocaRaton (2006).

[24] Paul D. R., Newman S.: Polyolefin blends: Rheology,melt mixing, and applications in polymer blends. Aca-demic Press, New York (1978).

[25] D’Orazio L., Mancarella C., Martuscelli E., Polato F.:Polypropylene/ethylene-co-propylene blends: Influ-ence of molecular structure and composition of EPRon melt rheology, morphology and impact propertiesof injection-moulded samples. Polymer, 32, 1186–1194 (1991).DOI: 10.1016/0032-3861(91)90220-D

[26] Prut E. V., Erina N. A., Karger-Kocsis J., MedintsevaT. I.: Effects of blend composition and dynamic vul-canization on the morphology and dynamic viscoelas-tic properties of PP/EPDM blends. Journal of AppliedPolymer Science, 109, 1212–1220 (2008).DOI: 10.1002/app.28158

[27] Da Silva A. L. N., Tavares M. I. B., Politano D. P.,Coutinho F. M. B., Rocha M. C. G.: Journal ofApplied Polymer Science, 66, 2005–2014 (1997).DOI: 10.1002/(SICI)1097-4628(19971205)66:10

<2005::AID-APP17>3.0.CO;2-2[28] Bielinski D. M., Slusaraki L., Wlochowicz A., Slusar-

czyk C., Douillard A.: Some aspect of isotacticpolypropylene crystallization in an ethylene-propy-lene-diene rubber matrix. Polymer International, 44,161–173 (1997).DOI: 10.1002/(SICI)1097-0126(199710)44:2<161

::AID-PI837>3.0.CO;2-T[29] Martuscelli E., Silvestre C., Abat G.: Morphology,

crystallization and melting behavior of films of isotac-tic polypropylene blended with ethylene-propylenecopolymers and polyisobutylene. Polymer, 23, 229–237 (1982).DOI: 10.1016/0032-3861(82)90306-8

[30] Boyce M. C., Yeh O., Socrate S., Kear K., Shaw K.:Micromechanics of cyclic softening in thermoplasticvulcanizates. Journal of Mechanics and Physics ofSolids, 49, 1343–1360 (2001).DOI: 10.1016/S0022-5096(00)00077-6

[31] Soliman M., Dijk M. V., Es M. V., Shulmeister V.:Deformation mechanism of thermoplastic vulcanisatesinvestigated by combined FTIR and stress-strainmeasurements. in ‘Proceedings of ANTEC ‘99, NewYork, USA’ vol 2, 1947–1954 (1999).

209

Babu et al. – eXPRESS Polymer Letters Vol.4, No.4 (2010) 197–209